All of the fuel tested in this program changed in octane rating as the pressure of the altitude chamber was decreased. In general, the fuels that were the most sensitive to altitude changes were those with the high9s.t se nsitivity to engil~e severity, as defined by the dIffer ences between thClr Motor Method and Researc~l Method rati!lgs. ~his is to be expected, as the hJgh er compreSSIOn ratIOs used at lower chamber pressures. increase the engine severity by raising the compr eSSIOn temperatures.It has been shown that decreased inlet temperature and decreased speed shift altitude ratings by the Motor Method toward their sea-level values, but at different rates for various fu els. Although th e effects of spark advance on this method were not investigated, a change in this factor might prove advantageous. Larger carburetor venturis have proved useful, and permit ratings to be maintained at about their ea-Ievel values up to altitudes to 3,000 to 4 000 ft. It is possible that the Motor Method rating~ are capable of being equalized throughout the altitude range by usc of a suitable combina tion of changed conditions.The Research Method, on the other hand do cs not lend itself well to modifications of this' type. fhe spark advanc.e is already nearly the optimum, mlet temperature lS at about the minimum that will assure adequate vaporization of the fuel the volumetric efficiency is improved very little by use of a larger venturi, and the engine speed is already very low.It appears, therefore, that the surest way of making the engine severity equal to that a t sea level, regardless of the altitude, is to sup ercharge the inlet and thro~t~e the exhau~t, thereby simulating sea-level conclttIOns. For thls purpose, a small single-sta,ge centrifugal compressor would probably suffice. The ail' would enter the blower through a standard humidity-controll ing ice towel', and thence through the s tandard air h eater to a pressurized carburetor. It is possible that an after-cooler may also be necessary to k eep the inlet temperature below 125 0 F in the R esearch Method at the lower chamber pressures..This investigation was carri ed out in cooperation wlth the Octane Correlation Advisory Committee of the Division of R efining of the American Petroleum Institute. Members of this committee arc: Bruno R. Siegel, chairman, Frank C. Burk, Carl E. Habermann, J . E. Taylor; associate members are Harold M. Trimble, H. R. Stacey , John M. Snell, Afton D. Puckett, and Hudson W. K ellogg. Appreciation is expressed to the committee members for planning th e scope of the work and their expert advice in car~'ying it o~t. Cred.it is also due th e following for thCJ.r work ltl operatmg the engines and alti tude eqll1pment: IV. C. Lacey, Myron C. Wolfe, Jame Creep tests we; e made at l10 0 , 250 0 , and. 300 0 F on a.nn~aled oxyge n-free high-conductIvIty copper .. lhe rate of 10adll1g to the ultImate had a slgn Ifl cant eff ect on t h e amount of plastic ext e nsIOn and t hereby affected the creep behavior. The strain rate durin g the so-calle...